the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 May 2023
Evaluation of modelled climatologies of O3, CO, water vapour and NOy in the upper troposphere – lower stratosphere using regular in situ observations by passenger aircraft
Abstract. Evaluating the global chemistry models in the upper troposphere – lower stratosphere (UTLS) is an important step toward a further understanding of its chemical composition. The latter is regularly sampled through in situ measurements based on passenger aircraft, in the framework of the In-service Aircraft for a Global Observing System (IAGOS) research infrastructure. This study focuses on the comparison of the IAGOS measurements in ozone, carbon monoxide (CO), nitrogen reactive species (NOy) and water vapour, with a 25-year simulation output from the LMDZ-OR-INCA chemistry-climate model. For this purpose, we present and apply an extension of the Interpol-IAGOS software that projects the IAGOS data onto any model grid, in order to derive a gridded IAGOS product and a masked model product that are directly comparable to one another. Climatologies are calculated in the upper troposphere (UT) and in the lower stratosphere (LS) separately, but also in the UTLS as a whole, as a demonstration for the models that do not sort out the physical variables necessary to distinguish between the UT and the LS. In the northern extratropics, the comparison in the UTLS layer suggests that the geographical distribution in the tropopause height is well reproduced by the model. In the separated layers, the model simulates well the water vapour climatologies in the UT, and the ozone climatologies in the LS. The opposite biases in CO in both UT and LS suggest that the cross-tropopause transport is overestimated. The NOy observations highlight the difficulty of the model in parameterizing the lightning emissions. In the tropics, the upper-tropospheric climatologies are remarkably well simulated for water vapour, as the observed CO peaks due to biomass burning in the most convective systems, and the ozone latitudinal variations. Ozone is more sensitive to lightning emissions than to biomass burning emissions, whereas the CO sensitivity to biomass burning emissions strongly depends on the location and on the season. Through this evaluation, the present study demonstrates that the Interpol-IAGOS software is a tool facilitating the assessment of the global model simulations in the UTLS, potentially useful for any modelling experiment involving chemistry-climate and chemistry-transport models.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2945 KB)
-
Supplement
(3035 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2945 KB) - Metadata XML
-
Supplement
(3035 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-572', Anonymous Referee #1, 21 Jun 2023
The manuscript presents a comparison of O3, CO , NOy and H2O from a nudged simulation of the LMDZ-OR-INCA chemistry-climate model with a set of long-running observations of these trace gases from instrumentation carried on-board commercial aircraft. The comparison is made over the upper troposphere / lower stratosphere region from observations made during cruise-level flights using the Interpol-IAGOS software package described in an earlier paper. To account for the strong discontinuities around the tropopause, separate comparisons can be made for the lower stratosphere and the upper troposphere. In addition, the comparison to observations are made for two sensitivity simulations: one with emissions from biomass burning turned off, and one with emissions from lightning turned off.
The observations shows many of the large scale features of the distribution of these trace gases associated with, for example, the differing height of the tropopause between the tropics and the mid-latitudes, the seasonal and regional nature of biomass burning emissions and the impact of monsoon circulations. The comparison with the model shows the model has an ability to reproduce many of these features, with the most significant and widespread bias being for CO in the lower stratosphere. The paper is generally well organized, though the use of particular phrases or words makes comprehension challenging in a few places. I have tried to point these out in the minor comments.
My only significant concern is the treatment of the vertical height coordinate when comparing the aircraft observations to the model. From Section 2.3.1, I can understand how the IAGOS observations are horizontally gridded. And I can understand how the IAGOS data is binned for the version of the data separated into the upper troposphere and the lower stratosphere. With most (or all?) of the data from cruise altitude, what I am missing is how variations in the cruise altitude are treated. At line 197 it is stated ‘The climatologies here refer to nearly horizontal maps derived from partial columns in the cruise altitudes.’ Are the point measurements assumed to be representative of a certain vertical range to allow a partial column to be calculated? Or are all the different aircraft observations in a particular month kept on their individual altitude points and these are combined to produce a vertically integrated (or vertically averaged) quantity at each grid point? Or maybe the relatively small variations in altitude are ignored and all data is assumed to be on an average cruise altitude? It is particularly important for the comparison in the lower stratosphere, which shows such strong vertical gradients in CO and ozone. Is it sufficient just to classify observations at ‘lower stratosphere’ without taking into account the distance to the tropopause?
Minor comments:
Line 2 – The use of ‘the latter’ in ‘The latter is regularly…’ does not work here because the preceding sentence does not present two options.
Line 15 – is there a word missing in ‘as [are] the observed CO peaks due to biomass burning…’
Line 139 – Does ‘In this study, the LMDZ GCM surface zonal and meridional wind components are nudged…’ mean that only the surface winds (lowest model layer) are nudged and the rest of the atmosphere is allowed to freely evolve?
Line 168 – ‘precising’ might be better as ‘denoting’
Line 179 – what is meant by ‘if it is not adjacent to the 2 PVU isosurface’? What comes between a particular grid point and the 2PVU isosurface so that it is not adjacent?
Lines 173 – 195 – A more general comment on Section 2.3.2: There is no discussion of how PV is used to classify the IAGOS measurements. I assume reanalysis data is interpolated in space and time to each IAGOS measurement and PV is calculated with the same exclusion of the transition layer?
Line 188 – I think ‘Consequently, it is worth figuring out that…’ would be better as ‘Consequently, it is worth keeping in mind that…’ or maybe ‘worth remembering…’?
Line 219 – I am not sure what is meant by ‘In the tropics, the threshold is adapted to the seasons duration by applying a cross product.’ A cross product is usually an operation on vectors in linear algebra.
Line 349 – For ‘we chose to draw the mean ratio’, in place of ‘draw’ that can have different meanings, could I suggest ‘plot’ or ‘display’?
Lines 353 – 354 – I think it is a bit of a jump to suggest that biomass burning and lighting are realistically distributed because the simulations without these important sources has a poorer comparison with observations.
Lines 354 – 366 – This section jumps around a lot across the different panels of Figure 6 and is difficult to follow. For example,
First, the comparison between the different runs shows a better correlation in the reference simulation in the UT, possibly suggesting that the effects from biomass burning and lightning emissions on ozone production are realistically distributed in space. As expected, no change is observed in the LS for this metric, since the higher amounts of ozone in the LS increase the NOx threshold necessary to trigger a net ozone production (e.g. Hegglin et al., 2006).’
From the second sentence it is not clear what metric is being discussed when it is stated ‘no change is observed in the LS for this metric’.
Lines 367 – 368 – ‘the shorter and lesser NOy sampling does not lead to strong differences.’ might be clearer as ‘the shorter period of time and sparser measurements of NOy does not lead to strong differences.’
Lines 373 – 375 – I would agree with the statement ‘It is likely that the influence of biomass burning on the LS is overestimated because of an excessive exchange between the troposphere and the stratosphere.’ The change in bias in CO in the LS is quite surprising.
Line 384 – I can deduce what is meant by ‘barycentre’ but it is not a correct word. ‘mean pressure of the measurements’ maybe?Citation: https://doi.org/10.5194/egusphere-2023-572-RC1 - AC1: 'Reply on RC1', Yann Cohen, 06 Oct 2023
-
RC2: 'Comment on egusphere-2023-572', Anonymous Referee #2, 06 Jul 2023
This manuscript provides climatological chemical composition information regarding the upper troposphere (UT) and lower stratosphere (LS), based on a fairly long-term compilation of in situ aircraft data (from IAGOS, as described in some detail), focusing on a gridded/averaged version of these data sets for O3, CO, NOy, and H2O, and a comparison versus a similarly gridded chemical model (LMDZ-OR-INCA). The comparisons entail annual mean mapped comparisons and seasonal comparisons as well, scatter plots, along with Taylor diagrams to highlight certain aspects of the model runs, which include simulations with no biomass burning or with no lightning NOx impacts; this helps to evaluate certain contribution factors regarding these climatologies in the UTLS. Although Taylor diagrams are not always liked by everyone, and there are a fair number of such plots, I found that for this purpose, they do add some useful information regarding the overall model characteristics and top-level fits to the data. The separation between the UT and LS regions, and the "all data regions together (UTLS)" was found to be useful as well. This manuscript is appropriate for publication in ACP, after some minor adjustments, minor, overall.
My main request (or query, at least) has to do with whether slightly more information could be mentioned regarding the water vapour lower stratospheric data and related comparisons, for both the data and model implications. The authors point out that the drier conditions in this region mean that the IAGOS H2O sensor is likely to have a wet bias, as mentioned on page 16 before section 3.2.1, and the filter they apply to try to ameliorate this issue leads to a low bias, still, for the model versus the IAGOS H2O data in this region (midlatitude lower stratosphere). Is it clear (at all) whether the model or the data might be more "at fault" for the LS discrepancies? If the data set is really expected to come with a high bias, can the morphology of the differences still be of some use or are we just stuck not being able to deduce much, except in the UT region, where the comparisons are viewed as among the best (between O3, CO, NOy, and H2O). Maybe there is a way to make some comparisons with other data sets (e.g., ballonborne data sets from Boulder, or elsewhere, and/or with satellite data sets, see for example the Read et al., 2022 article in AMT, on UT H2O intercomparisons, if this might be useful). I will accept that this probably goes beyond the current goals/expectations for this manuscript, if no further comments can be currently made about the disagreements between the model and the data for H2O. However, since there might be indications of an overestimated tropopause/stratosphere exchange, with high model CO biases, for example, in the lower stratosphere, would this not also lead to larger model values of H2O in the LS? Yet, the model clearly underestimates the IAGOS H2O values there. It is unfortunate if nothing more robust can even be mentioned (or possibly even speculated about), but if this is the case, this will need to await better measurements and/or validation exercises for the IAGOS LS H2O data. The authors could still (try to) mention something slightly more informative, even if they have been leary not to do so, so far.
Ignoring the above, especially if it seems too challenging of a topic, given the current state of the LS H2O data, I do have a number of quite minor wording comments, with a few suggestions for clarifications, as listed below.
After these matters are considered/answered, I see no reason not to recommend this manuscript for publication, as the comparisons offer some nice insights to the community regarding these species inn the UTLS, and they highlight mostly good comparisons with the model that is presented. As the authors point out, it would indeed be useful for future studies to pursue multi-model comparisons with IAGOS data, in order to enable some better discrimination among the various models regarding their representation of various processes (e.g., transport, convection, biomass burning, lightning NOx sources, and ultimately longer-term variations in the UTLS).
Mostly technical and a few minor comments/questions:- Line 1, change "the global" to "global"
- L2, I suggest: toward an improved understanding of the chemical composition in this region. This composition is regularly sampled...
- L7, I would suggest adding "(sub-sampled)" after "masked", for clarity.
- L12, I suggest: There are opposite model CO biases in the UT and the LS, which suggests that...
- L14,15, need to reword the 2nd half of this sentence so it can be properly understood. Please clarify the meaning ("and the ozone latitudinal variations" is not tied to the rest of the sentence, and the CO part is also not tied that well, and does not directly follow from a good simulation for water vapour). Maybe use a semi-colon after "water vapour" and then write what you want to convey for the 2nd part of the sentence.
- L17, depends on location and season. The present study demonstrates that...
- L18, assessment of global model simulations...
- L19, "chemistry climate or chemistry transport models."
- L23, with varying strength
- L26, ozone absorbs most of the energetic ultraviolet...
- L27, the air's oxidizing capacity
- L29, regarding the formation and life cycle of cirrus clouds, whose large radiative forcing still carry a large uncertainty...
- L31, thus increasing the CH4 lifetime.
- L33, classified as essential climate variables
- L34, NOx gets converted back and forth into its reservoir species ...
- L37, chemical species and their radiative forcings
- L40, uncertainties in dynamical processes.
- L42, NOy also provides information
- L44, chemical destruction, NOy can also...
- L45, mass origins... CO, on one hand, and O3 and NOy on the other hand, are more ...
- L48, The assessment of CCM or CTM simulations relies on comparisons with
- L49, few observations are suitted for diagnosing... and few can account for UTLS vertical
- L55, ", but they are too sparse..."
- L57, provide frequent and large-scale
- L58, programs have highlighted large-scale feautres since the 1970s; these programs include TROZ..., GASP..., and more recently, NOXAR..., [all?] with an observation period of four years or less.
- L62, Since more than two decades ago, the In-service... has provided...
- L68, took advantage of the whole IAGOS database.
- L80, of the impact of lightning and...
- L81, the present study goes further into the development and application ...
- L84, O3, CO, but also H2O...
- L89, biomass burning to the modelled...
- L90, account for differences in the definitions of seasons and in the mean tropopause altitude.
- L95, Its predecessors, MOZAIC... and CARIBIC..., relied on the same principle.
- L101, Since the merging...respective databases are referred to as ...
- L102, hereafter, with an approach validated by...
- L104, measured with an ultraviolet ...
- L112, sensor ranges from 5 to 70% RHL...
- L118, the case of the frost-point...
- L123, you need to define the acronym "ORCHIDEE" here.
- L130, aerosols scheme includes a total of 123 tracers and 22 aerosol tracers.
- L133, please provide a reference (or more) to clarify this statement regarding past comparisons.
- L137, mostly dealing with chlorine and bromine chemistry, along with 66 gas-phase ...
- L150, consistent with the annual...
- L176, which ensures that the UT and the LS are sufficiently isolated...
- L179, please clarify "if it is not adjacent to the 2 PVU isosurface"; I suppose you meant the PV value is lower than a particular threshold (what is this value, or what do you mean by adjacent here?).
- L180, is less than 400 hPa.
- L181, the model's ability to reproduce the chemical...
- L183, mean value is greater (less) than
- L188, it is worth noting that the water...
- L189, other measurements, for which there is no such filter.
- L191, This study presents quasi-horizontal maps and quantifies the mean ...
- L193, Consequently, part of this software...
- L197, partial columns at cruise altitudes. Please define better (somewhere) hos the partial columns are derived and how you arrive at the units (say, parts per billion) based on "columns"; see also Figures 1-4, for example...ensuring that the vertical grid is briefly defined (without having the reader trying to find this in an earlier reference). Please clarify how broad the "typical" regions are in the UT and in the LS (separately), regarding the IAGOS measurements used in this work; maybe pointing to Figure 7 is sufficient, but please help the reader make this connection, if possible.
- L206, change "meridian" to "meridional" [and at several other places in the manuscript].
- L218, a factor of 4 for the...
- L227, change "this couple of metrics is defined as" to "these two metrics are defined as:"
- L236, We use these metrics to evaluate the reference simulation.
- L244, these Figures 1-4 should have been better explained in terms of how you obtain average VMR values based on "column measurements" mentioned earlier. Even if a past reference explains this better, providing a brief summary here would be useful.
- L249, sampling relative to the tropopause...geographically, as a result of tropopause and cruise altitude variations.
- L255, the minimum in the western equatorial ...
- L258, like the good model...
- L260, characterized by a smaller geographical variability...
- L262, moderate positive CO bias...
- L263, NOy is characterized by discrepancies...
- L266, a noticeable minimum East of Central America.
- L267, NOy tends to be overestimated
- L268, NOy is underestimated. meridian --> meridional
- L270, maximum in H2O [or CO or?] above the most convective regions...
- L271, Atlantic Ocean, as well as the collocated ozone minimum. This [H2O or O3?] feature is due to...
- L290, visible changes in the MNMB or in the correlation.
- L291, but at most, it can be interpreted as an upper limit.
- L299, For a given species, we note that there are high correlations [between what and what? model and data?] in the layer where the mixing ratios are at a maximum (LS for ozone,...).
- Table 2 is good, although you probably do not need that many digits for MBNB and FGE (-0.086 could be written as -0.09 for example) [it is not critical to change this however, but it is not needed or significant enough in my view; same for Table A1].
- L335, water vapour mixing ratios at low latitudes...
- L340, nitrogen have poorer scores, with lower correlation coefficients and an underestimated geographical variability.
- L345, The Taylor diagrams in Fig. 6 present a synthesis of ...
- L355, is it not also true that the longer chemical lifetime for ozone in the LS implies more of a dynamical control than a chemical one [this is not exactly the same as just stating that O3 abundances are larger in this region]?
- L356, This is consistent with the fact...
- L359, please clarify which r value corresponds to which run.
- L361, change "lightnings" to "lightning"
- L363, 'layer, such as too much convection.' This overestimated tropospheric influence could be consistent (it sems) with the overestimated H2O amounts, it seems to me; you may want to comment about this.
- L364, more important in the run without LNOx.... In the "no LNOx" run, model ozone has a significant negative bias (from -15 to -20% ... in annual means), and model NOy has a small bias (...), while model CO is increased and shows a 10-50% positive bias. [is this how one should write/read this sentence?]
It also seems that an issue with model transport (or mixing) could affect the model NOy values [but not clear if this would be consistent with too much LS model H2O].
- L372, I suggest rewording as follows [or clarifying]: "In the run with no biomass burning, we observe decreases in CO, and the annual model CO bias changes from -5% to -15% in the UT, from 30 to 15% in the LS, and from 15 to 0% in the UTLS. Surprisingly, the impact of biomass burning is not negligible in the LS, especially in the summer.
- L374, again, an excessive net model transport between troposphere and stratosphere goes along with high values of H2O...
- L376, "This suggests that this season maximizes..."
- L384, measurements' barycentre . They can be associated with significant...
- L387, are thus difficult to interpret [if this is what you mean]
- L392, Their small number of occurrences indicates that seasonal mean...
- L393, which provides more confidence regarding the representativeness of the NOy measurements in the context of the whole ozone measurement period.
- L401, This is consistent with the peak... [and can probably delete "calculated"]
- L407, above the South American tropics...
- L408, association with the start...
- L423, This is consistent with enhanced ozone...
- L424, as shown by Barret et al. (2016), and as suggested by the important seasonal...
[but what does "important seasonal O3/CO ratio mean? please clarify, does this mean "large seasonal O3/CO ratio variations highlighted by Cohen et al. (2018)?]. I would also break up this long sentence: "; this was also confirmed by measurements from the ...HALO during the HALO-ESMVal campaign...
- L428, stratospheric intrusions to the northern side
- L430, They linked such important photochemical activity with a combination of uplifted precursors...
- L434, Good consistency between... obsrvations is visible for ozone...
- L436, well the H2O maximum at 5-10...
- L437, along the northern tropics...
- L440, has difficulties in reproducing the extremely...
- L453, above Africa from December to March and from June to October...
- L457, The simulated NOy profiles underestimate the observed meridional variability.
- L458, Above Africa, NOy is almost systematically underestimated by the model...hemisphere, but the NOy comparisons show a general consistency in the northern hemisphere.
- L459, an important positive NOy model bias during the Asian summer...
- L468, simulations show some strong variability.
- L469, lightning NOx in CO destruction...
- L473, notably between the opposite subtropics, probably reflecting an underestimated modeled [?] interhemispheric transport.
- L476, CO local maxima simulated...May are not visible
- L492, consistent with the literature...
- L496, is associated with local biomass..., , as [is?] a significant part of the peak
- L497, In contrast, the observed CO maximum during April-May...is rather associated with other sources.
- L500, two regions is mainly caused by biomass burning.
- L501, consistent with the literature.
- L503, reduces wildfires via enhanced precipitation.
- L505, This study presents an assessment of...
- L521, underestimated by the model in the UT, nd model CO shows a positive bias in...
- L522, overestimation in the model's extra-tropical cross-tropopause mixing [or net transport?]. Again, such a model feature might affect the large values of H2O (?).
- L523, This is likely linked...
- L525, tracks; aircraft can indeed play a significant role in the NOy levels.
- L531, I would also strongly recommend that you add a sentence (or two) here regarding the large values of modeled H2O in the LS.
- L534, able to accurately represent the mean...
- L538, this sentence is not clear enough; you quote 3 numbers (25, 30, 45 ppb CO) but the previous sentence seems to mention only two seasons (D-M and J-O) unless you need to add the convective season above South America (please specify which months), or clarify the 3 numbers and their relationship to the previous sentence more carefully.
- L539, you should probably (briefly) explain in this section the CO sink from lightning emissions, for readers that might not read the whole paper...
- L542, where enhanced CO is attributed to biomass burning peaks.
- L544, inconsistencies in model ozone and CO with respect to...
- L547, is likely to enhance the model skills for NOy and ozone.
- L549, daily output or the model monthly output
- L551, in the framework of the second phase
- L553, Other potential applications include ... scales and for seasonal scales, as well as interannual..., possibly also allowing for source apportionment regarding the observed features.- Figure 5 (3rd line in caption), fit described in the top-left corner... [rather than on the top-left corner]. Change "the amount of grid points" to "the number of grid points".
- Figure 6, the assessment of the yearly climatologies...
Citation: https://doi.org/10.5194/egusphere-2023-572-RC2 - AC2: 'Reply on RC2', Yann Cohen, 06 Oct 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-572', Anonymous Referee #1, 21 Jun 2023
The manuscript presents a comparison of O3, CO , NOy and H2O from a nudged simulation of the LMDZ-OR-INCA chemistry-climate model with a set of long-running observations of these trace gases from instrumentation carried on-board commercial aircraft. The comparison is made over the upper troposphere / lower stratosphere region from observations made during cruise-level flights using the Interpol-IAGOS software package described in an earlier paper. To account for the strong discontinuities around the tropopause, separate comparisons can be made for the lower stratosphere and the upper troposphere. In addition, the comparison to observations are made for two sensitivity simulations: one with emissions from biomass burning turned off, and one with emissions from lightning turned off.
The observations shows many of the large scale features of the distribution of these trace gases associated with, for example, the differing height of the tropopause between the tropics and the mid-latitudes, the seasonal and regional nature of biomass burning emissions and the impact of monsoon circulations. The comparison with the model shows the model has an ability to reproduce many of these features, with the most significant and widespread bias being for CO in the lower stratosphere. The paper is generally well organized, though the use of particular phrases or words makes comprehension challenging in a few places. I have tried to point these out in the minor comments.
My only significant concern is the treatment of the vertical height coordinate when comparing the aircraft observations to the model. From Section 2.3.1, I can understand how the IAGOS observations are horizontally gridded. And I can understand how the IAGOS data is binned for the version of the data separated into the upper troposphere and the lower stratosphere. With most (or all?) of the data from cruise altitude, what I am missing is how variations in the cruise altitude are treated. At line 197 it is stated ‘The climatologies here refer to nearly horizontal maps derived from partial columns in the cruise altitudes.’ Are the point measurements assumed to be representative of a certain vertical range to allow a partial column to be calculated? Or are all the different aircraft observations in a particular month kept on their individual altitude points and these are combined to produce a vertically integrated (or vertically averaged) quantity at each grid point? Or maybe the relatively small variations in altitude are ignored and all data is assumed to be on an average cruise altitude? It is particularly important for the comparison in the lower stratosphere, which shows such strong vertical gradients in CO and ozone. Is it sufficient just to classify observations at ‘lower stratosphere’ without taking into account the distance to the tropopause?
Minor comments:
Line 2 – The use of ‘the latter’ in ‘The latter is regularly…’ does not work here because the preceding sentence does not present two options.
Line 15 – is there a word missing in ‘as [are] the observed CO peaks due to biomass burning…’
Line 139 – Does ‘In this study, the LMDZ GCM surface zonal and meridional wind components are nudged…’ mean that only the surface winds (lowest model layer) are nudged and the rest of the atmosphere is allowed to freely evolve?
Line 168 – ‘precising’ might be better as ‘denoting’
Line 179 – what is meant by ‘if it is not adjacent to the 2 PVU isosurface’? What comes between a particular grid point and the 2PVU isosurface so that it is not adjacent?
Lines 173 – 195 – A more general comment on Section 2.3.2: There is no discussion of how PV is used to classify the IAGOS measurements. I assume reanalysis data is interpolated in space and time to each IAGOS measurement and PV is calculated with the same exclusion of the transition layer?
Line 188 – I think ‘Consequently, it is worth figuring out that…’ would be better as ‘Consequently, it is worth keeping in mind that…’ or maybe ‘worth remembering…’?
Line 219 – I am not sure what is meant by ‘In the tropics, the threshold is adapted to the seasons duration by applying a cross product.’ A cross product is usually an operation on vectors in linear algebra.
Line 349 – For ‘we chose to draw the mean ratio’, in place of ‘draw’ that can have different meanings, could I suggest ‘plot’ or ‘display’?
Lines 353 – 354 – I think it is a bit of a jump to suggest that biomass burning and lighting are realistically distributed because the simulations without these important sources has a poorer comparison with observations.
Lines 354 – 366 – This section jumps around a lot across the different panels of Figure 6 and is difficult to follow. For example,
First, the comparison between the different runs shows a better correlation in the reference simulation in the UT, possibly suggesting that the effects from biomass burning and lightning emissions on ozone production are realistically distributed in space. As expected, no change is observed in the LS for this metric, since the higher amounts of ozone in the LS increase the NOx threshold necessary to trigger a net ozone production (e.g. Hegglin et al., 2006).’
From the second sentence it is not clear what metric is being discussed when it is stated ‘no change is observed in the LS for this metric’.
Lines 367 – 368 – ‘the shorter and lesser NOy sampling does not lead to strong differences.’ might be clearer as ‘the shorter period of time and sparser measurements of NOy does not lead to strong differences.’
Lines 373 – 375 – I would agree with the statement ‘It is likely that the influence of biomass burning on the LS is overestimated because of an excessive exchange between the troposphere and the stratosphere.’ The change in bias in CO in the LS is quite surprising.
Line 384 – I can deduce what is meant by ‘barycentre’ but it is not a correct word. ‘mean pressure of the measurements’ maybe?Citation: https://doi.org/10.5194/egusphere-2023-572-RC1 - AC1: 'Reply on RC1', Yann Cohen, 06 Oct 2023
-
RC2: 'Comment on egusphere-2023-572', Anonymous Referee #2, 06 Jul 2023
This manuscript provides climatological chemical composition information regarding the upper troposphere (UT) and lower stratosphere (LS), based on a fairly long-term compilation of in situ aircraft data (from IAGOS, as described in some detail), focusing on a gridded/averaged version of these data sets for O3, CO, NOy, and H2O, and a comparison versus a similarly gridded chemical model (LMDZ-OR-INCA). The comparisons entail annual mean mapped comparisons and seasonal comparisons as well, scatter plots, along with Taylor diagrams to highlight certain aspects of the model runs, which include simulations with no biomass burning or with no lightning NOx impacts; this helps to evaluate certain contribution factors regarding these climatologies in the UTLS. Although Taylor diagrams are not always liked by everyone, and there are a fair number of such plots, I found that for this purpose, they do add some useful information regarding the overall model characteristics and top-level fits to the data. The separation between the UT and LS regions, and the "all data regions together (UTLS)" was found to be useful as well. This manuscript is appropriate for publication in ACP, after some minor adjustments, minor, overall.
My main request (or query, at least) has to do with whether slightly more information could be mentioned regarding the water vapour lower stratospheric data and related comparisons, for both the data and model implications. The authors point out that the drier conditions in this region mean that the IAGOS H2O sensor is likely to have a wet bias, as mentioned on page 16 before section 3.2.1, and the filter they apply to try to ameliorate this issue leads to a low bias, still, for the model versus the IAGOS H2O data in this region (midlatitude lower stratosphere). Is it clear (at all) whether the model or the data might be more "at fault" for the LS discrepancies? If the data set is really expected to come with a high bias, can the morphology of the differences still be of some use or are we just stuck not being able to deduce much, except in the UT region, where the comparisons are viewed as among the best (between O3, CO, NOy, and H2O). Maybe there is a way to make some comparisons with other data sets (e.g., ballonborne data sets from Boulder, or elsewhere, and/or with satellite data sets, see for example the Read et al., 2022 article in AMT, on UT H2O intercomparisons, if this might be useful). I will accept that this probably goes beyond the current goals/expectations for this manuscript, if no further comments can be currently made about the disagreements between the model and the data for H2O. However, since there might be indications of an overestimated tropopause/stratosphere exchange, with high model CO biases, for example, in the lower stratosphere, would this not also lead to larger model values of H2O in the LS? Yet, the model clearly underestimates the IAGOS H2O values there. It is unfortunate if nothing more robust can even be mentioned (or possibly even speculated about), but if this is the case, this will need to await better measurements and/or validation exercises for the IAGOS LS H2O data. The authors could still (try to) mention something slightly more informative, even if they have been leary not to do so, so far.
Ignoring the above, especially if it seems too challenging of a topic, given the current state of the LS H2O data, I do have a number of quite minor wording comments, with a few suggestions for clarifications, as listed below.
After these matters are considered/answered, I see no reason not to recommend this manuscript for publication, as the comparisons offer some nice insights to the community regarding these species inn the UTLS, and they highlight mostly good comparisons with the model that is presented. As the authors point out, it would indeed be useful for future studies to pursue multi-model comparisons with IAGOS data, in order to enable some better discrimination among the various models regarding their representation of various processes (e.g., transport, convection, biomass burning, lightning NOx sources, and ultimately longer-term variations in the UTLS).
Mostly technical and a few minor comments/questions:- Line 1, change "the global" to "global"
- L2, I suggest: toward an improved understanding of the chemical composition in this region. This composition is regularly sampled...
- L7, I would suggest adding "(sub-sampled)" after "masked", for clarity.
- L12, I suggest: There are opposite model CO biases in the UT and the LS, which suggests that...
- L14,15, need to reword the 2nd half of this sentence so it can be properly understood. Please clarify the meaning ("and the ozone latitudinal variations" is not tied to the rest of the sentence, and the CO part is also not tied that well, and does not directly follow from a good simulation for water vapour). Maybe use a semi-colon after "water vapour" and then write what you want to convey for the 2nd part of the sentence.
- L17, depends on location and season. The present study demonstrates that...
- L18, assessment of global model simulations...
- L19, "chemistry climate or chemistry transport models."
- L23, with varying strength
- L26, ozone absorbs most of the energetic ultraviolet...
- L27, the air's oxidizing capacity
- L29, regarding the formation and life cycle of cirrus clouds, whose large radiative forcing still carry a large uncertainty...
- L31, thus increasing the CH4 lifetime.
- L33, classified as essential climate variables
- L34, NOx gets converted back and forth into its reservoir species ...
- L37, chemical species and their radiative forcings
- L40, uncertainties in dynamical processes.
- L42, NOy also provides information
- L44, chemical destruction, NOy can also...
- L45, mass origins... CO, on one hand, and O3 and NOy on the other hand, are more ...
- L48, The assessment of CCM or CTM simulations relies on comparisons with
- L49, few observations are suitted for diagnosing... and few can account for UTLS vertical
- L55, ", but they are too sparse..."
- L57, provide frequent and large-scale
- L58, programs have highlighted large-scale feautres since the 1970s; these programs include TROZ..., GASP..., and more recently, NOXAR..., [all?] with an observation period of four years or less.
- L62, Since more than two decades ago, the In-service... has provided...
- L68, took advantage of the whole IAGOS database.
- L80, of the impact of lightning and...
- L81, the present study goes further into the development and application ...
- L84, O3, CO, but also H2O...
- L89, biomass burning to the modelled...
- L90, account for differences in the definitions of seasons and in the mean tropopause altitude.
- L95, Its predecessors, MOZAIC... and CARIBIC..., relied on the same principle.
- L101, Since the merging...respective databases are referred to as ...
- L102, hereafter, with an approach validated by...
- L104, measured with an ultraviolet ...
- L112, sensor ranges from 5 to 70% RHL...
- L118, the case of the frost-point...
- L123, you need to define the acronym "ORCHIDEE" here.
- L130, aerosols scheme includes a total of 123 tracers and 22 aerosol tracers.
- L133, please provide a reference (or more) to clarify this statement regarding past comparisons.
- L137, mostly dealing with chlorine and bromine chemistry, along with 66 gas-phase ...
- L150, consistent with the annual...
- L176, which ensures that the UT and the LS are sufficiently isolated...
- L179, please clarify "if it is not adjacent to the 2 PVU isosurface"; I suppose you meant the PV value is lower than a particular threshold (what is this value, or what do you mean by adjacent here?).
- L180, is less than 400 hPa.
- L181, the model's ability to reproduce the chemical...
- L183, mean value is greater (less) than
- L188, it is worth noting that the water...
- L189, other measurements, for which there is no such filter.
- L191, This study presents quasi-horizontal maps and quantifies the mean ...
- L193, Consequently, part of this software...
- L197, partial columns at cruise altitudes. Please define better (somewhere) hos the partial columns are derived and how you arrive at the units (say, parts per billion) based on "columns"; see also Figures 1-4, for example...ensuring that the vertical grid is briefly defined (without having the reader trying to find this in an earlier reference). Please clarify how broad the "typical" regions are in the UT and in the LS (separately), regarding the IAGOS measurements used in this work; maybe pointing to Figure 7 is sufficient, but please help the reader make this connection, if possible.
- L206, change "meridian" to "meridional" [and at several other places in the manuscript].
- L218, a factor of 4 for the...
- L227, change "this couple of metrics is defined as" to "these two metrics are defined as:"
- L236, We use these metrics to evaluate the reference simulation.
- L244, these Figures 1-4 should have been better explained in terms of how you obtain average VMR values based on "column measurements" mentioned earlier. Even if a past reference explains this better, providing a brief summary here would be useful.
- L249, sampling relative to the tropopause...geographically, as a result of tropopause and cruise altitude variations.
- L255, the minimum in the western equatorial ...
- L258, like the good model...
- L260, characterized by a smaller geographical variability...
- L262, moderate positive CO bias...
- L263, NOy is characterized by discrepancies...
- L266, a noticeable minimum East of Central America.
- L267, NOy tends to be overestimated
- L268, NOy is underestimated. meridian --> meridional
- L270, maximum in H2O [or CO or?] above the most convective regions...
- L271, Atlantic Ocean, as well as the collocated ozone minimum. This [H2O or O3?] feature is due to...
- L290, visible changes in the MNMB or in the correlation.
- L291, but at most, it can be interpreted as an upper limit.
- L299, For a given species, we note that there are high correlations [between what and what? model and data?] in the layer where the mixing ratios are at a maximum (LS for ozone,...).
- Table 2 is good, although you probably do not need that many digits for MBNB and FGE (-0.086 could be written as -0.09 for example) [it is not critical to change this however, but it is not needed or significant enough in my view; same for Table A1].
- L335, water vapour mixing ratios at low latitudes...
- L340, nitrogen have poorer scores, with lower correlation coefficients and an underestimated geographical variability.
- L345, The Taylor diagrams in Fig. 6 present a synthesis of ...
- L355, is it not also true that the longer chemical lifetime for ozone in the LS implies more of a dynamical control than a chemical one [this is not exactly the same as just stating that O3 abundances are larger in this region]?
- L356, This is consistent with the fact...
- L359, please clarify which r value corresponds to which run.
- L361, change "lightnings" to "lightning"
- L363, 'layer, such as too much convection.' This overestimated tropospheric influence could be consistent (it sems) with the overestimated H2O amounts, it seems to me; you may want to comment about this.
- L364, more important in the run without LNOx.... In the "no LNOx" run, model ozone has a significant negative bias (from -15 to -20% ... in annual means), and model NOy has a small bias (...), while model CO is increased and shows a 10-50% positive bias. [is this how one should write/read this sentence?]
It also seems that an issue with model transport (or mixing) could affect the model NOy values [but not clear if this would be consistent with too much LS model H2O].
- L372, I suggest rewording as follows [or clarifying]: "In the run with no biomass burning, we observe decreases in CO, and the annual model CO bias changes from -5% to -15% in the UT, from 30 to 15% in the LS, and from 15 to 0% in the UTLS. Surprisingly, the impact of biomass burning is not negligible in the LS, especially in the summer.
- L374, again, an excessive net model transport between troposphere and stratosphere goes along with high values of H2O...
- L376, "This suggests that this season maximizes..."
- L384, measurements' barycentre . They can be associated with significant...
- L387, are thus difficult to interpret [if this is what you mean]
- L392, Their small number of occurrences indicates that seasonal mean...
- L393, which provides more confidence regarding the representativeness of the NOy measurements in the context of the whole ozone measurement period.
- L401, This is consistent with the peak... [and can probably delete "calculated"]
- L407, above the South American tropics...
- L408, association with the start...
- L423, This is consistent with enhanced ozone...
- L424, as shown by Barret et al. (2016), and as suggested by the important seasonal...
[but what does "important seasonal O3/CO ratio mean? please clarify, does this mean "large seasonal O3/CO ratio variations highlighted by Cohen et al. (2018)?]. I would also break up this long sentence: "; this was also confirmed by measurements from the ...HALO during the HALO-ESMVal campaign...
- L428, stratospheric intrusions to the northern side
- L430, They linked such important photochemical activity with a combination of uplifted precursors...
- L434, Good consistency between... obsrvations is visible for ozone...
- L436, well the H2O maximum at 5-10...
- L437, along the northern tropics...
- L440, has difficulties in reproducing the extremely...
- L453, above Africa from December to March and from June to October...
- L457, The simulated NOy profiles underestimate the observed meridional variability.
- L458, Above Africa, NOy is almost systematically underestimated by the model...hemisphere, but the NOy comparisons show a general consistency in the northern hemisphere.
- L459, an important positive NOy model bias during the Asian summer...
- L468, simulations show some strong variability.
- L469, lightning NOx in CO destruction...
- L473, notably between the opposite subtropics, probably reflecting an underestimated modeled [?] interhemispheric transport.
- L476, CO local maxima simulated...May are not visible
- L492, consistent with the literature...
- L496, is associated with local biomass..., , as [is?] a significant part of the peak
- L497, In contrast, the observed CO maximum during April-May...is rather associated with other sources.
- L500, two regions is mainly caused by biomass burning.
- L501, consistent with the literature.
- L503, reduces wildfires via enhanced precipitation.
- L505, This study presents an assessment of...
- L521, underestimated by the model in the UT, nd model CO shows a positive bias in...
- L522, overestimation in the model's extra-tropical cross-tropopause mixing [or net transport?]. Again, such a model feature might affect the large values of H2O (?).
- L523, This is likely linked...
- L525, tracks; aircraft can indeed play a significant role in the NOy levels.
- L531, I would also strongly recommend that you add a sentence (or two) here regarding the large values of modeled H2O in the LS.
- L534, able to accurately represent the mean...
- L538, this sentence is not clear enough; you quote 3 numbers (25, 30, 45 ppb CO) but the previous sentence seems to mention only two seasons (D-M and J-O) unless you need to add the convective season above South America (please specify which months), or clarify the 3 numbers and their relationship to the previous sentence more carefully.
- L539, you should probably (briefly) explain in this section the CO sink from lightning emissions, for readers that might not read the whole paper...
- L542, where enhanced CO is attributed to biomass burning peaks.
- L544, inconsistencies in model ozone and CO with respect to...
- L547, is likely to enhance the model skills for NOy and ozone.
- L549, daily output or the model monthly output
- L551, in the framework of the second phase
- L553, Other potential applications include ... scales and for seasonal scales, as well as interannual..., possibly also allowing for source apportionment regarding the observed features.- Figure 5 (3rd line in caption), fit described in the top-left corner... [rather than on the top-left corner]. Change "the amount of grid points" to "the number of grid points".
- Figure 6, the assessment of the yearly climatologies...
Citation: https://doi.org/10.5194/egusphere-2023-572-RC2 - AC2: 'Reply on RC2', Yann Cohen, 06 Oct 2023
Peer review completion
Journal article(s) based on this preprint
Data sets
Interpol-IAGOS Yann Cohen, Virginie Marécal, Béatrice Josse, Valérie Thouret https://doi.org/10.25326/81
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 298 | 141 | 20 | 459 | 46 | 9 | 10 |
- HTML: 298
- PDF: 141
- XML: 20
- Total: 459
- Supplement: 46
- BibTeX: 9
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Didier Hauglustaine
Bastien Sauvage
Susanne Rohs
Patrick Konjari
Ulrich Bundke
Andreas Petzold
Valérie Thouret
Andreas Zahn
Helmut Ziereis
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2945 KB) - Metadata XML
-
Supplement
(3035 KB) - BibTeX
- EndNote
- Final revised paper